Integrated circuits are often formed from different processes depending on the desired functionality. Multiple chips of one function are generally constructed on a single wafer and then cut apart to create a single die or “chip.” In other instances, the multiple chips can have multiple functions and include die partitioning and multiple functional die integration. The chips have inputs and outputs (“I/O”) that need to be electrically connected to I/O connections of other chips. For example, a processing chip may need to be connected to a memory chip to properly perform a function. One way to electrically join two chips is to use an active or passive device known as a “bridge.” A bridge provides a pathway from an I/O connection of one chip to an I/O connection of another chip. The bridge can be a silicon die with connecting routes that are passive or have an active function. However, using a bridge to provide the connections often increases the manufacturing complexity of the semiconductor device and may, therefore, also limit who can manufacture the semiconductor device. The foregoing is often true for bridges formed in substrates which can limit the process to foundries (manufacturers of chips) as opposed to Outsourced Assembly and Test (OSAT) facilities (limited to wafer level manufacturing). Some wafer level processes that form redistribution layers (RDL) after bridge and chip placement risk having to discard the expensive chips due to low RDL yields.
Thus, the inventors have provided an improved method and apparatus for forming a wafer-level die bridge.